Parkinson’s disease determinants, prediction and gene-environment interac- tions: a UK Biobank study

Introduction Early identification of individuals at high-risk of Parkinson’s disease (PD) may allow for the use of disease modifying treatments. We aimed to define those at risk of incident PD through environmental and genetic risk factors, comorbidities and prodromal symptoms of PD. Methods We identified incident cases of PD (n=1276) in the UK Biobank and unmatched controls without a PD diagnosis (n=500,406). Odds ratios for a range of determinants were calculated. A non-genetic model and polygenic risk score (PRS) were constructed. Basic gene-environment interactions were explored. Results Strong evidence of association (FDR < 0.05) was found between PD and family history of PD or dementia, not smoking, low alcohol consumption, depression, daytime somnolence, epilepsy and earlier menarche. Individuals in the top PRS decile had an increased risk of PD compared to those in the lowest decline and high scores were associated with an earlier age of diagnosis. There was evidence for a strong interaction between diabetes and PRS. Discussion We report the novel finding of a strong association between earlier menarche and increased PD. Analysis of gene-environment interactions found strong evidence that the effect of diabetes is modified by genetic risk. Replication and understanding the mechanism of this observation is an important next step. d.belete@nhs.net